REAL TIME IMPLEMENTATION OF A BLIND BEAMFORMER FOR SUBBAND SPEECH ENHANCEMENT USING KURTOSIS MAXIMIZATION
Benny Sällberg, Mikael Swartling, Nedelko Grbić, Ingvar Claesson · Lund University Publications (Lund University) · 2006
This paper presents a real time implementation of a blind beamformer for subband speech enhancement. The beamformer adaptivelymaximizes the statistical kurtosis measure of the beamformer’soutput signal. Speech carries high kurtosis and noiseoften exhibit lower kurtosis. Hence, maximization of the outputsignal’s kurtosis enhances speech, in general. The implementationis carried out on a novel framework for real time audio processingin MATLAB and uses low latency ASIO sound cards.The implementation is evaluated using recorded signals and thespeech is enhanced approximately 10 dB by the proposed approachwith perceptually low speech distortion.